Tag Archives: neural network

574–585 S. Polyanskikh, I. Arinicheva, I. Arinichev and G. Volkova
Autoencoders for semantic segmentation of rice fungal diseases
Abstract |
Full text PDF (937 KB)

Autoencoders for semantic segmentation of rice fungal diseases

S. Polyanskikh¹, I. Arinicheva², I. Arinichev³* and G. Volkova⁴

¹Plarium Inc., 75/1 Uralskaya Str., RU350001 Krasnodar, Russia
²Kuban State Agrarian University named after I.T. Trubilin, 13 Kalinina Str., RU350044 Krasnodar, Russia
³Kuban State University, 149 Stavropolskaya Str., RU350040 Krasnodar, Russia
⁴All-Russian Research Institute of Biological Plant Protection, 1 VNIIBZR Str., RU350039, Krasnodar, Russia
*Correspondence: iarinichev@gmail.com

Abstract:

In the article, the authors examine the possibility of automatic localization of rice fungal infections using modern methods of computer vision. The authors consider a new approach based on the use of autoencoders – special neural network architectures. This approach makes it possible to detect areas on rice leaves affected by a particular disease. The authors demonstrate that the autoencoder can be trained to remove affected areas from the image. In some cases, this allows one to clearly highlight the affected area by comparing the resulting image with the original one. Therefore, modern architectures of convolutional autoencoders provide quite acceptable visual quality of detection.

Key words:

, , ,




1368–1379 D. Nemeikšytė and V. Osadčuks
Design of modified movement planning system as a component of an intelligent planning system for robot manipulator
Abstract |
Full text PDF (994 KB)

Design of modified movement planning system as a component of an intelligent planning system for robot manipulator

D. Nemeikšytė and V. Osadčuks*

Latvia University of Life Sciences and Technologies, Faculty of Engineering, 5 J. Cakstes blvd., LV-3001 Jelgava, Latvia
*Correspondence: nemeiksyte.daiva@llu.lv, vitalijs.osadcuks@llu.lv

Abstract:

Different fields of industry and in-service support widely use robots, mechatronic and robotic technology systems in their activities. This is related to growing functionalities that result from using more advanced control systems the development of which is based on available achievements in the technical measures of computing. Therefore, the subject of study in this article was movement of a robot manipulator in using a fuzzy logics and neural network, and the goal of the study was to develop methods for designing combined intelligent planning and control systems for robot-manipulator movement in static dynamic environments based on the combined use of fuzzy logic apparatus and artificial neural networks to reduce the possibility of robot-manipulator’s joints colliding into unknown obstacles located in its operating area. Based on this, the robot arm model has been developed after calculating in the article the missing parameters of the experimental robot manipulator in order to analyze the peculiarities of using the fuzzy logics device as well as the specifics and challenges of using neural network. As a result of the study performed in the article, significant data were obtained based on which a method was offered for an intelligent system for planning robot manipulator movement in static environment using a fuzzy blocks, which was characterized by the use of neural network corresponding each block, and localization of each solution to the task of planning robot manipulator movement in each specific situation, which enables to improve the accuracy and efficiency of movement planning.

Key words:

, , , ,




1993–2004 A. Nemeikšis and V. Osadčuks
Development of intelligent system of mobile robot movement planning in unknown dynamic environment by means of multi-agent system
Abstract |
Full text PDF (753 KB)

Development of intelligent system of mobile robot movement planning in unknown dynamic environment by means of multi-agent system

A. Nemeikšis and V. Osadčuks*

Latvia University of Life Sciences and Technologies, Faculty of Engineering, 5 J. Cakstes blvd., Jelgava LV-3001, Latvia
*Correspondence: nemeiksis.andrius@llu.lv, vitalijs.osadcuks@llu.lv

Abstract:

Through the ages the world has conceived the projects which are aimed at creating diverse models of robots that would be beneficial for exploration of different dangerous surfaces where human participation is excluded. Therefore, the main task of the study of this article is to develop the researches, the object of which is mobile robot movement in unfamiliar environment, based on multi agent apparatus system and neural networks. The aim of the research is to develop methods for creating intellectual systems for planning mobile robot movement in unfamiliar environment applying the methods of multi agent apparatus and neural networks ensuring the robot executes the planned and adjusted on the way safe trajectory in an environment with unknown obstacles. Accordingly, the entire study of the article is based on a two-stage process. The first stage involves determination of distance between the robot and the obstacles in its operating area as well as classification of the possible location of obstacles, based on the information received from distance sensors, using the model of multilayer neural networks. During the second stage bypassing obstacles, wall tracking, movement-to-destination as well as speed management agents are developed. As the result of the study, a method was suggested for creating neural network model for classification of environment into agents and their consistent switching, which, according to the classification table compiled, involves all the possible locations of obstacles occurring on the robot’s movement trajectory and allows reducing the number of unfamiliar environment situations that are necessary to identify.

Key words:

, , , ,